Skip to content

The Census and Statistical Sampling

The partisan wrangling over whether or not to allow the census bureau to supplement the traditional head-count method (‘actual enumeration’) with statistically-derived estimates for the expected under-count in the 2000 census was enlightening only in the way it illustrated our leaders’ ability to miss the point.

This letter was submitted to the San Jose Mercury News and the San Francisco Chronicle; it was published in the Mercury News on 31 January 1999.

26 January 1999

News reports and editorials on this week’s Supreme Court decision concerning statistical sampling and the census once again paint the Democratic party as the champion of the poor and downtrodden, wrestling righteously against those evil Republicans who are concerned only with political advantage. How, we wonder, can anyone in good conscience oppose the principle that the allocation of representation should fairly reflect the population, including, perhaps especially, those unfortunates that the census-by-enumeration cannot reach?

It is a compelling argument. Unfortunately, it is camouflage, based on the faulty assumption that to count them and allocate representation for them will make a difference. It will not.

It’s a pretty safe bet that those who vote are already counted by the census; conversely, it’s also all but a sure thing that estimating the numbers of those who are not currently counted — and who don’t currently vote — will not suddenly enfranchise them. Counted or not, they will still lead marginal lives on the fringes of society. Counted or not, they will still be transient, will still avoid ‘official’ entanglements. Counted or not, they still won’t vote, and therefore will not, in any real sense, be ‘represented’.

In short, this fight is not about providing representation to those who are not represented; it is about providing yet more representation to those who already have it. Make no mistake about this: if we, as the result of statistical sampling, allocate more congressional districts in places, like large cities, where we assume the uncounted reside, all we will succeed in doing is handing more electoral power to the existing electorate in those places. The uncounted disenfranchised will become the counted disenfranchised and the middle- and upper-class voters of those districts will elect more of the same.

That, of course, is what the Democratic party hopes for. If either they or the Republicans really cared about enfranchising those uncounted masses, they would figure out how to bring them into the mainstream of society, where the counting, even by enumeration, would take care of itself.

© Copyright 1999, 2005, Augustus P. Lowell

Leave a Reply